One of the most challenging steps of a simulation study is the output data analysis, since it provides the information necessary to make decision regarding project objectives. As a matter of fact, a very common practice is to carry out an arbitrary number of simulation runs of arbitrary length and then to perform inference on the simulation results, treating them as the “true” model output.The aim of this paper is to propose an innovative methodology to investigate simulation outputs in order to identify “sustainable target” for the system and express it as a point value rather than as a confidence interval. For this purpose, a decision support tool is then proposed, it takes as input the process output data of observed samples and provides as result the achievable targets for each couple of replication number and replication length by using the classical statistical concepts of type-I and type-II error.
Target detection for simulated systems / Perrica, Giuseppe; Goldoni, Gabriele; Grassi, Andrea; Fantuzzi, Cesare. - STAMPA. - (2009), pp. 557-564. (Intervento presentato al convegno MITIP2009, Modern Information Technology in the Innovation Processes of the Industrial Enterprises tenutosi a Bergamo, Italy nel October 15-16, 2009).
Target detection for simulated systems
PERRICA, Giuseppe;GOLDONI, Gabriele;GRASSI, Andrea;FANTUZZI, Cesare
2009
Abstract
One of the most challenging steps of a simulation study is the output data analysis, since it provides the information necessary to make decision regarding project objectives. As a matter of fact, a very common practice is to carry out an arbitrary number of simulation runs of arbitrary length and then to perform inference on the simulation results, treating them as the “true” model output.The aim of this paper is to propose an innovative methodology to investigate simulation outputs in order to identify “sustainable target” for the system and express it as a point value rather than as a confidence interval. For this purpose, a decision support tool is then proposed, it takes as input the process output data of observed samples and provides as result the achievable targets for each couple of replication number and replication length by using the classical statistical concepts of type-I and type-II error.Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris